 And then welcome to the Cube Pod episode 37. I'm John Furrier, Dave Vellante. Dave, we're on episode 37. I just got back from Chicago at KubeCon. You are in Miami for a Cisco event and a big IBM analyst events going on. Incredible week of AI action. It's, again, generative AI all the time every day. Every company is here. This is the, I mean, to me, the topic dominating the conversation continues to be generative AI and the impact. People are afraid of it. It's changing the paradigm shifts of all businesses. And this week, there was a couple of similar moments going on this week that are worth kind of highlighting. I mean, besides the economy, some of the things we'll get to in more of the analysis section is that you had two things that happened here this week that I think really point to the radical shift of AI that we've been talking about on the Cube Pod for a long time. Open AI had their dev day on this week, this Monday. It was pretty incredible. And yesterday, the AI pin was launched which was a moonshot project that is essentially a very Apple-esque product. And it was developed by former Apple iPhone creatives and engineers creating essentially a device, a wearable device that is, remind me of the Google Glass Conversations, a wearable device that's a phone powered by T-Mobile. And it's a really elegant device that takes pictures, picks up sounds, it's almost like a body, camera, personal assistant. You can get gesture, augmented gesturing, just incredible. And it's campaigned by a company called Humane. AI, it's a case study for wearable future. And- Amazing how fast this came out. It's really- It's revolutionary. Well, a lot of people were like, well, Steve Jobs would be wrong. He never would approve that script. But it was both the open AI and the pin was a significant thing because you had two modern launches. One was a keynote. And you pointed out that strategy article from Ben Thompson, what I thought was right on. It's very Apple-esque. I was actually commenting that on the Cube, the CubeCon was people were cheering. It felt like it was only their first event for open AI and people were cheering. This is a cult-like status. Open AI has so much momentum and they're doing a great job. And I was very impressed and inspired by Sam Altman and the team. It was a tight keynote packed with great messaging. They maximized their time. And then the AI pin that was launched yesterday was perfectly orchestrated like in a great video. It was just an elegant device, pimping up the fact that it's beautiful, the next Apple kind of vibe. And people were really hammering it. And I thought that was not fair because to me, it was witnessing like a moonshot. A moonshot is a term that they say for the big audacious ideas, that big goals you wanna go for something in the moon is a big idea. And they spent a lot of time building this device, both from an operating system standpoint, the connectivity, the capabilities. It's elegant looking. It's not an eyesore device like the Google Glass was. It's more complicated. It's a moonshot device, will it fail? Maybe, but it was damn cool, I thought. Now the privacy people are freaking out because it's like, dude, we're talking about cameras on your body. It's like body cams. We were joking years ago. Remember we said people gonna have body cams and streaming their own lives and then bring into the cube. And we had this vision of like, hey, everyone can walk around, reinvent and we'll just run a kinesis, a pipeline into the cloud and stream it to the cube. Well, guess what? That shit might happen. So this is like a very big moonshot. And then obviously KubeCon, huge implications there again. And infrastructure show, Genevieve AI was front and center, ethical concerns around privacy, security, end to end. And then just a lot of news. You have some good news and some bad news. You saw a lot of still getting funding rounds are down significantly. And again, on the business side, every single company is looking for their AI tailor, right, the AI clothing manufacturer. Everyone wants to change their clothes and put on their AI wares. We are AI, IBM was a classic example. I'm watching Rob Thomas up there. It's almost like they're putting the clothes back on again and like putting on the new fashion. Every single company is going through the following motions. Update your talk, update your talk track, do this. I mean, you were there. I mean, from the far, it looked weak. But that's why it was IBM. Well, was it, it wasn't public, right? Was it? There was tons of stuff going on in public. I saw a bunch of people. Yeah, there's stuff on social. Yeah. It was far from weak. It was actually really impressive. That's funny. Why do you say that it looked weak? What did you see that suggested it was weak? And the slides were like unbelievably detailed. Of course, people can share the slides. It felt like, well, it felt like everyone, I'm skeptical now of the AI because it's like everyone's trying to say, look, we had AI before it was AI. Okay. And I know Rob Thomas because we quoted him in a very positive way in the cube saying he had AI before AI, information architecture. And, you know, Rob Thomas is a great guy. We love Rob's, not getting on him. Just saying is that it just feels like IBM and how they're talking about it. It feels like they're just late to the party and they're trying too hard to be cool. Well, I would, I would say. And, you know, I don't like the Watson name. I think that's, that's- No, so this is, that's the topic of my breaking analysis today. Cause I mean, I've kind of been a skeptic here, but then when we did that show with the guys in the IBM storage group, there's the data, the storage summit that we had out in Palo Alto. I interviewed Vincent Shu, who is like an alpha geek in IBM. And he was telling me, cause I didn't go to think this year where they announced Watson X. So I kind of didn't pay attention to it. I'm like, okay, that's like same wine, new bottle. And, but then when I talked to Vincent this summer, I was like, holy shit. Are you shipping this? He goes, yes. We announced in May and then they've announced all these other modules, AI and then now they got governance coming in, in, in December. And they've got this really robust roadmap. They, I think IBM finally got it right with Watson. I mean, Watson 1.0 was just a fail. You know, my, what I say in my breaking analysis today is, you know, the Silicon Valley mantra fail fast. Well, Watson 1.0 didn't, it took like a decade to fail. But, but Watson 2.0 is, is actually the balls. I mean, it's really, really good. They have a robust stack from Silicon all the way through the analytics up to the, the ISVs and the stack, their own SAS. And they, they've got AI chips. They've got partnerships. Obviously, hugging face is a big partner. I would tell you, I was really impressed and here's why. So I have been so critical of IBM's inability in the 2010s to actually turn R and D into product. IBM's got all these like gorgeous research facilities. Like we were just at the Thomas J. Watson research facility in Yorktown Heights. And I had been there like years and years ago when I was at IDC, back when IBM was, you know, king. And I've been to the Alamedan. I've been to a number of IBM research facilities. You've been, right? They're, they're unbelievably gorgeous. We were, we covered, we covered IBM's AI before anybody else. I mean, everyone else on the market was like behind on the curve. Yeah, but I know. But so, but they were trying to make Watson. It's like we talked to Jeff Jonas about this. They were trying to make Watson do things that it wasn't intended to do. Like, you know, solve cancer and, you know, climb trees. And so, but so what was really interesting was the cultural change that I see in IBM from the standpoint of getting, going from research to product and commercializing and dropping it to the bottom line income statement. And I asked IBM, I asked Dario Gill, like, you know, like what's changed? I want some specifics on this. And his answer was so good. He's like, look it, we used to IBM research. They used to run around. This was really interesting. They used to say each research, whether it was, you know, Almaden or Yorktown Heights or Japan, you have to be known for something. What are you going to be known for? Ooh, ooh, I'll take database. Ooh, I'll take supercomputing or whatever it is. And they had all these bespoke projects going on. So we've done away with that. And now it's like they have the resources are pooled and the product, the engineering teams are much closer to the research. And so they've really shifted the culture. And then the other thing, and it was like six or seven other really key dimensions that I won't go into in the interest of time. But then Rob Thomas said something that was really interesting. He said, I want to say it was like three months ago. He said, if you're in product, your job is product. It used to be, if you're in product, you did like, you know, one third product management, one third product marketing, one third go to market. And he said, no more. If you're in product, you're a hundred percent in product. So much more focused. We don't need you and go to market. And he kind of hinted that that sort of pissed some people off but he was like, I don't give a shit. He didn't, this is my words, not his. You don't like it leave. We're going to focus. Product management is product management. Product marketing is product marketing. Go to market is go to market. So focus. And I just, the cultural shift to me was palpable. And then when you combine it with what they're actually delivering, the last thing I'll say is, if you look at the data in the spending data, the ETR data, IBM in April of 2023, the month before they announced Watson X was like below the line. I mean, it looked kind of like pretty shitty. And in the October survey, it's shot up and surpassed Oracle. So the point is, it seems like when somebody announces something that's substantive, they get spending momentum. We saw it with obviously chat GPT and well, Microsoft and open AI. We definitely saw it with Google when they announced in May, Vertex AI. We're seeing it with IBM. And I think we're going to see it big time with Bedrock that just went GA. And so I was impressed, John. I have to say. Well, I'm glad that was good overview on IBM. Thanks for sharing. I'm a little bit skeptical more on that. We'll see how they do. But generally, IBM, like every other company out there, has to come out and this is a feature, not a bug. They have to come out and essentially reboot their image. Okay, I mean, IBM's got huge momentum in the marketplace. So I'm not surprised to see some of the increase of accelerators and questions. Can they sustain it? Every company has to come out and say, we're a generative AI company. After the open AI dev days clear that the developers are on board from the cheering and all the reactions from the content and they ship product, people are using it. I was playing with it. And then look at KubeCon. We just came back from Chicago. I was there for four days and this is the open source conference for cloud native, the cloud open source community, which is all embracing generative AI. And we had multiple analytics sessions on that. And it's clear that you can throw all the messaging you want and all the grandstanding and all the people cheering for the companies like IBM and others who are going to say, look at us, look at us. At the end of the day, the open source is a big part of the innovation. If you look at the power law research that we put out, that Kube research team formerly Wikibon put out, you see the power law is getting massive traction at our work there. And why is it getting traction? Is because the open AI dev day points specifically to why traditional incumbents are struggling, okay? IBM struggling, Amazon struggling, Cisco might be struggling. Maybe they're going to have something to come out because they have to essentially, as leaders, adopt it as fast as possible, slow it down or speed up. They got to either slow down the potential new entrance so that they can catch up. And as we pointed out, the people who have, the bigger companies who implement AI get better fast. So AI will favor the big companies. So I'm expecting AWS to tool up huge. I'm expecting IBM to tool up really huge. Cisco to tool up huge with AI because they have to, if they don't, they'll be behind. And they know that. So it's not like surprise. Every single other company, even in open source, Dave, are bringing AI to the table. Okay, experimentation and in production and in the open source world, huge conversation around keeping it open, not closing it down, okay? And then debate shift to, can the big models like open AI, which is actually, is ironically being called a proprietary model even though it's open by the name and they crawl the web, they're offering Turbo, GPT for Turbo and it has your own GPT. So it's got to get traction. I mean, inference is the new web app was Tim Hawkins' line at the keynote yesterday, closing day keynote. Tim Hawkins was one of the inventors of Kubernetes, Cube alumni. Hey, I thought that was compelling. And Dustin Kirkland, I was talking about that, Dave, he said the new web app is inference, meaning GPT is gonna be the new web app. You're gonna see more GPT, custom GPT than sub-stacks or YouTube videos. Now people are gonna start automating their workflows, a little bit of RPM meets AI kind of thing there, Dave. So generative AI is gonna get massive traction in open source. So I thought the open AI announcements were noteworthy. They were shipping products. They increased their context window, which is huge for things for ingesting articles, just to put it in perspective, 128 bit context windows, like a 300 page book. That's huge. Okay, that means all of our transcripts now from these videos can be sucked into the GPT turbo. And that's gonna be a dream. The retrieval side of it, I think it's gonna be interesting. The vector databases are providing great value for companies. And I think that's gonna continue to be a feature, not a company, but it's gonna be part of data stores. So I think that's gonna coexist. So open AI was a huge win. The pebble, the pebble, called pebble, the pin, AI pin, that's the side show. Before you get off open AI, what would impress me was the integrations that they're doing. Should you remember, early on it was like, okay, we can go out to Expedia, we can go to American Airlines, we can book a hotel, whatever it is. But those were really clunky integrations. And I don't think they're shipping this yet, but what they showed was like this seamless integration. Where the developer doesn't have to invoke anything, it's just services that are there that are invoked by natural language. That is gonna be, I mean, I'm amazed at the pace at which they're moving. You know what I'm talking about with that integrations? Yeah, the partner integrations, open AI. Yeah, but specifically, the example I'm using is travel. Book me a trip to Milan. And I wanna stay at this type of hotel, and this is what I wanna pay. And before it was really clunky to be able to do that. The developer would have to actually invoke the service. I mean, you have to actually do stuff, like clicks. Whereas now it's just gonna happen through natural language. And that is unbelievably powerful. And I think the neural network aspect of AI is gonna be a big part of the infrastructure. And again, I think we said it a few pods ago that we were pontificating like a new AI system is gonna come out. I think the open AI is gonna leverage their investment in the large language model and all the results from that training and make it an inference engine and let people bring their data sets in. That's what they announced with their private partner initiative. And by the way, they're only gonna be doing that as a professional service for only a handful of people. But the idea of having that figured out is gonna be a huge win. Private LLMs is gonna be a big deal. And that came up in, at KubeCon, Oracle has a partner called Cohere, a CUBE alumni, I've been on there, founders of there. They're working with OCI to do private LLMs. This is a huge feature. And that's the first thing. Oracle Microsoft expanding their partnership. Did you see that? Yeah, exactly. That was pretty amazing. So I mean. That's super cloud. Those guys are building a super cloud despite what insect Ray said, poo poo. Well, yeah, you lost me at Oracle. Remember he said that at super cloud one? You lost me at Oracle. I think Oracle's at an all-time high, insect. Busted my balls. It's all about the stock price, Dave. You can engineer stock price in many different ways. So, you know. One way is to print cash, John. Oracles. Can I just say one other thing before we get off of, I wanna correct myself because on a breaking analysis and maybe on the Q pod, I said that other than Nvidia, Microsoft was the only company that, you know, laying its Gen AI chops on the table. They said they had 300 basis point tailwind, incremental tailwind from Gen AI for Azure, which equates to probably 350, 400 million. I said, those are the only two, Nvidia and Microsoft open AI. It's not true. Arvind Krishna, I missed it and I went back and checked this morning. He said we did low hundreds of millions. And of course, one of the analysts rightly so was my first question. Yeah, how much that was consulting and services. And he was honest. He said about half was consulting and the other half was software. So you're talking about north of a hundred million in revenue from IBM. And I think Amazon, my assumptions are in cloud that Amazon is gonna show an uplift in growth this quarter. They had a little uplift last quarter, as you know, as I track this stuff. I think it's going to be a couple of hundred basis points at least from Gen AI this quarter now that bedrock is GA. And I think after reinvent going into 2024, it's going to be even a bigger boost. And it's all about acceleration and feature delivery. And you know, Charles Fitzie, his latest says, you know, he's trying to Microsoft or AWS to give up. No way they're building like mega LLMs and they're not going to give up. They're going to dominate in my opinion. Well, I'm going to go see Adam Sileski next Friday. So a week from today, I'll be in Seattle sitting down for a exclusive interview with the CEO of AWS. And I'm going to ask him about this stuff. And I think, you know, they're getting hammered through a leaked story around them building Olympus, which is their huge language model. You know, by the way, Fitzie's taking pot shots because he's Microsoft. And he's just, Why are they getting hammered? That's like, isn't that like two trillion parameters? Well, I mean, people, first of all, a lot of the commentary coming from Charles Fitzgerald, mainly because he's Microsoft dogmatically a Microsoft guy and he's got it actually running with Amazon. So he's, he's, he's, he's, but his, his real, I think the contribution is his CapEx report, which was quoted in financial times, but I thought that was solid, big story there. Amazon's not going away. Again, this is going to be a long game. He, it's an easy target piece. I mean, it's not hard to attack Amazon right now with some of the moves they're making. Cause you know, it looks like they're groping and we'll see what comes out of re-invent. And you know, you pointed out Dave this week on, I think it was on a Twitter, we were texting back and forth, we were on our chat group. Jassy's involved now, right? So, and then we predicted that on the cube pod, by the way. Yeah, yeah. He runs AI. It reports to him. He's not screwed around. Remember, remember, Andy Jassy and Adam Sileski were a very strong team as a management team. Most people don't know this. A lot of the other analysts that are kind of new on the scene, we see him out there. They haven't been around for a very long time. They're kind of new and they're kind of, the shallow hot takes out there running around saying, hey, look at, we know what you're talking about. They haven't been around. Like they haven't even been around to even kind of get the power dynamics. But, and then they're crapping on Amazon. Jassy and Sileski worked together, even before he went to Tablo to get some CEO chops, which he then boomerang back. That's what they call it when they come back. And that's why Jassy embraced them. And by the way, Jassy pretty much will embrace anyone that comes back. He told me that personally. He's not against that at all, unless it's, they burned the bridge. But Jassy is like, if I had to put Adam and Jassy, they're two different, it's like two people in the same sports team. One has got certain skill player, and the other one could be the quarterback, one's a receiver, one's a runner. So they're definitely complimentary, but they're different skill sets. Andy goes deep on product stuff, loves to talk about product, and he gets studies hard. He's really in the books. Adam is really much on the more of a holistic, big thinker around strategy, sales, and marketing more heavily on that side. So we look at their strengths and weaknesses. I see them teaming up on this. Well, I'm going to try to confirm that more when I sit down with them with Adam, but they're definitely working together. And Adam's not a lightweight either. He knows product, he knows what's going on. The question is, can they pull a rabbit out of the hat for reinvent, right Dave? So we're calling it the battle for cloud supremacy on our live stream. We're going to run that week. We're doing a special program called super cloud five battle for AI supremacy, which is really is. I mean, Microsoft's going to have their ignite conference next week, Dave. That's going to be massive. You don't think they're going to come into that conference, guns blaring, post, open AI. Because remember, Satya Nutella came on stage. He flew down to Silicon Valley for this event. I wouldn't he. I saw him at like 10 billion reasons. He's at the great event with Reid Hoffman. So, you know, they're going to really rock. They're going to rock the rock the house with their cloud show. Did you? Oh, I'm Amazon's going to do really what people don't understand about it. Look at first of all, I totally agree. Amazon's playing catch up here. They got caught flat footed, but everybody got caught flat footed. But Amazon is always used to being the horse on the on the lead, you know. And they like to be out front. They like to have the most features. They like to have people following them, right? And so because they're closer to the customers, right? They're obsessed with the customer. But in this case, you know, somebody created something. It was like the iPhone moment. It's like, oh, shit, let's go. But people forget that Amazon's always had a strong AI play, ML play with SageMaker. And they've got the Silicon chops and they're ahead in Silicon. People can argue they're catching up, but it's hard to catch up in Silicon. So I think they're going to do very, very well in generative AI. I think all three cloud players are going to do well in generative AI. We're seeing, you definitely see Anthropic popping up in the spending momentum. Like I mentioned, I mentioned IBM. I'm really curious to see, you know, if they'll keep moving up, one quarter does not make a trend. Actually, there's an interest in Oracle. So because they've got, you know, they've got some AI chops, people forget, but the cloud guys are well, well ahead because they have the tooling. But to your other point, there's gonna be a lot of stuff done on private AI on-prem about, I would estimate that a little over half of the Lama 2 downloads are going, you know, on-prem experimentation, let's say, based on my sources. And I got to ask you, did you get my questions for Adam? No, I did not. I spent a bunch of questions. No, I did not. Damn. You didn't get those? No, I didn't. I sent some killer questions. You asked for, like, you know, our input. I know. You got to find those. It's kind of a plane like a midnight last night. So I'll look at it today. Okay, make sure you find those because I put a lot of thought into it. I mean, it was kind of real time, but I'll dig them out if you don't have them. Are you gonna see Jassy? Well, last year, you had some great questions. You and Rob Hoef and Paul Gillan and Mark Albuson really made that last year's exclusive. I use half my questions were from you. And then the other third was from Rob and the editorial team. I think it only got three of mine in. You guys crushed it. I think, you know, Timmy, I mean, again, I don't want to rattle on Amazon. We talk a lot about Amazon. Timmy, I'm more excited by the open AI and the power law that we put out there. I'm telling you right now, if you're listening to this and you haven't seen our power law research, you have to go check it out. It's getting massive traction. We're adding more onto it, but that is what's happening. If you have a long tail developing, you're gonna have a big torso developing, these specialized models are coming out. It's clear that in open source, these long context LLMs as these content, when context windows open up, again, very nerdy kind of in the weeds here, Dave, but the super important with these vectorizations and these retrievals, vector embeds, you now have the ability to remember things differently. You brought this up earlier about these use cases. I want to go travel to a hotel. Well, how's it gonna know that you like a certain thing? Well, they might have checked your email and have to remember that. How are they gonna remember that? Well, the technology isn't keyword based, it's math based. So what's happening is when you have a larger context window, like it understanding your preferences or understanding what's in the model, it's got to remember it and then create these similarities, these math equations, whether it's Bayesian and what all the technology use. The bigger the context window, the recall specifics drop. You need more tokens. And so, summarizing tasks, do writing blog posts, the stuff that people see now is trivial and that looks magical, but it kind of hits a point where it's not good, right? And then that's where the next open source challenge is people who are really talking about, for example, the Lama Index, okay? They've been doing a lot of stuff. Jerry Lew over there, he's gonna come on the queue, by the way, for our SuperCloud Five. He's the CEO of Lama Index. They're on GitHub. They got a great guy. He's got benchmarks out. They were pre-testing GPT-4 128K context recalls, okay? Significant drop-off, okay? And so on performance. So it's not as good as people think. So you and I were talking about before the podcast is RAG or Retrieval Augmented Generation going to be challenged? Howie Xu and I were talking on Twitter about this right after the announcement that we thought that with the OpenAI's GPT-4 Turbo, that retrieval and RAG might get impacted. So as it turns out, a lot of the data is proving that that's not the case. It just increases the need for better retrieval augmentation because the context window is taking in more context. So it's a whole new science here for what I call search or retrieval, memory, context. Completely different paradigm shift, Dave, than Google search or keyword search. I've been pimping up, as you know, the cube AI and I didn't realize it's now, you don't have to, it's not private beta anymore. It's like kind of public beta, right? You can go to thecubeai.com and just play with it. And so but, because there's a lot of confusion about RAG, same thing. People are like, oh, RAG's dead now. And I'm like, oh, I don't think so. And so to- Well, we got to update the cube. I just got a text from Savannah Peterson. I love the AI tool, except for this one made me giggle. What were people saying about KubeCon in Chicago? We don't have an answer for that. Well, yeah, because it happened yesterday. Yeah, I mean, we haven't, you know, it's close to real time as possible. We got to get the assets in there and sync it up. Look, I am digging the AI because I kind of am squinting through those short term things we fix, but it's, it got good answers, Dave. I'm telling you, I think it's, so ask it what Rob Thomas thinks about AI and IA. And the answer will be off the charts. Ask it about Cisco's strategy. What's Cisco's AI strategy? What's Dell's AI strategy? I mean, that's awesome. And the point I'm making is we found, it's not so much the hallucinations. That's not our problem. The issue for us is gaps in the data. In other words, you know, like when Dave Donatelli came in with Riverbed, you know, we said, all right, what's Riverbed's known for? And it was like really outdated because there's not updated data in there. So you got to get the updated data in there. And we're not crawling the web. We have our own corpus. And then I just wanted to make a comment on the power law. The power law, the post that we did last week was really well received. And then this guy, this data engineer, he's a Python developer, came into my LinkedIn and said, interesting data, not the best visualization. So he's busting my chops on my shitty PowerPoint skills. So I got to work on that. Tell them we're having a hackathon. Come on in. Yeah, come on in. Open source that graphic. Top prize is $5,000. Let's go, you know. Hey, make it pretty. Well, I'm working on a cube graphic too. It's funny. I've been using Dolly to get images. It's pretty amazing on a cube research kind of presentation of scoring companies by their relevance because you and I were talking about that the other day and I'm like working on this algorithm to test. Because everyone asked me, you know, back up why you think IBM's not strong. And I have a reason. I'm going to have, I'm going to report on that. When I say not strong, I mean like not the leader. And maybe it's because I haven't gotten the word out but I got to get more up to date when you and I should circle back on that. Well, the cloud guys are leading. I mean, there's no doubt about that. I mean, Microsoft and OpenAI are leading. No question. Google is number two in my opinion. Amazon is I think number three just because of its SageMaker chops. And it's what it's got, what it just went GA with. And I think Amazon's going to, I don't know if it's going to overtake Microsoft and OpenAI, that's going to take some work. But I think it's going to be, it's going to be a really good tailwind for their business. And so, oh, I'm obsessed. I just don't think, I don't think their innovation is. Wait, wait, wait, wait, IBM's not. They're not, they're not there with these guys but they have the technical foundation and they have differentiation. I get it. I get it. I just have a different opinion. And they get a stack and because of Red Hat, they have a hybrid story. It's really about AI and hybrid cloud for them. And so now they have to prove it and get the word out. I mean, they're not, they don't have the momentum of the cloud guys. Don't get me wrong. But it seems to me, the other big takeaway was, no longer is the consulting tail wagging the dog. They are getting back to their product roots. And it's been a long time since I've been dominated. What products are dominating? No, they're not dominating. I wouldn't say they're dominating. They haven't dominated since mainframes in the ThinkPad. So I have, wait, wait, wait, wait, wait, wait, wait, but what, wait, let me just. Oh, then the ThinkPad. Then say something. Okay, I think, I mean, I don't want to get too far over our skis on Quantum because Quantum is like this crapshoot. But if Quantum actually becomes real, IBM is going to be in a good position. But it's going to take probably a decade for them to show that. But I think their AI, I don't think they're going to lead, I don't think they're going to lead like be number one in AI. I don't think that's going to happen. I think the cloud guys will. But I think IBM is going to do very, very well with Arvin's achievable strategy of hybrid cloud and AI and consulting to drive a lot of industry specific to the power law. And IBM's going to throw off a lot of cash doing that. Well, I'm looking forward to the products that actually will sell in an innovative way versus waiting for Quantum. But my perspective is simple. Disruption factor, how is the company disrupting the current market? Seems like a me too. Adoption rate, you mentioned some momentum. How's it a me too? How's it a me too? They've got a full stack of AI. They've announced, they announced in May, they announced in July, what's an X for AI? What's an X for data? They got a governance module coming out. They got silicon. How is that a me too? Well, it's their AI story. Everyone's got their AI story. The question is, how does that disrupt the, does that technology disrupt the current market? Yes or no? Does that technology disrupt the current market? Yes or no? I think. I have a high bar. I think IBM. I mean, I have a different bar than you do. I think IBM. I'm not seeing it even in the charts other than recycling their existing AI story. Oh my God, no, not at all. Not at all. Well, I'll get the data. It's truly, it's truly. Well, look at it. I'm just making it. But wait. Okay, Dave, Dave, Dave. I'm not litigating IBM. You don't need to answer for IBM. No, but to your question, I'm not. But to your question about, is it disrupting, I don't think that's the metric. Is it disrupting the market? I don't think IBM is going to, they're not the disruptor. They're the incumbent. They're going to sell to their existing client base. They're going to sell to mainframe customers. They're not going to, open AI is disrupting the market. Microsoft pulled a massive judo movement is disrupting the market. Is Amazon's AI disrupting the market? I mean, if IBM doesn't really think they want to be disrupted in the market so their customers can get the benefit of their technology, then I don't think they got the right talent in there. But that's. Is Amazon disrupting the market in AI? Not yet. So they would Google. Google is Google. I think has disruption that going on. I would give them more disruption score than then maybe I don't understand what disruption means. What does disruption mean? So when I put together my is disruption mean you're following open AI and Microsoft? No, a disruptive enabler. You come in with something that's new, not that's going to change in great value, enable value proposition. So, so Dave, I'm talking about something of scoring a company. So I look at a copy of saying, are they disrupted? Do they have a disruptive factor? Okay, some have more progressive disruption, messaging or position and others do. Some take a different approach. Some take different approaches. You know this, you're an analyst, right? So it's like, preach to the choir. And then you say, okay, their product, does it have that disrupting enablement capability in markets like today, not quantum future market today, the AI market customers want what they're expecting to see. User interface changes using data, creating business value. Okay, that is the market today. Game disruption. That's business that business value. Disruption doesn't equal business value. Of course it does. You change old, you make the old better. Or you take- But the definition of business value is are you driving economic value for your customer? It's not. And software does that, Dave. Software does that. But you're saying disruption is a litmus test of value? It's one of them, yes, it is. Okay, let me ask you a question. Can you create value and not be a disruptor? Depends what the market is, Dave. It's like, because I think Cisco, because that's the other thing that was at Cisco last week. I think Cisco has got a great business, but I don't think Cisco's a disruptor. But Cisco's got an awesome business and they're going to deliver value for customers. If you're in a mature market that has no need for disruption because it's, say, a steady state, then you have different dynamics in how you manage that market. Their growth rates are different. The expectations are different. This is how you manage that market's different. That's kind of how it works, right? Yeah, but every market's going to get disrupted by AI. In growing markets where there's a shift happening, there's value changes, new entrants come in, there's opportunities, things change. For example, you might have, you mentioned travel. Maybe there's a travel data model that comes out from AI that creates some disruptive positive enablement that creates more value, that creates an opportunity to sell another product, a net new product. That's a disruptive enabler. That's a disruption factor. Okay, that's like having the right product at the right time in a growing TAM market. Do they have adoption? Do they have a product that gets adopted? What's their ecosystem look like? If they have partner channels, they got to go direct, indirect, investment in R&D. How well are they doing? I call that the innovation score. I like these scoring. I like your scoring frame. Hey, I'm just saying. I like it. You can apply it into, it is kind of a score. So you say, okay, Amazon, you go back to 2015. They would check all the boxes. Yeah, but all I'm saying is that to me, the question, no question, but the real issue for the incumbents is can they take advantage of the disruption? Can they move fast enough to take advantage of the disruption? They don't have to invent. I think you might be thinking that you have to invent disruption. You can write it. I thought you were saying that they're not the disruptor. And I'm like, I agree with you. IBM is not the disruptor. Cisco is not the disruptor. Dell is not the disruptor. Well, a disruption factor for Dell or Cisco would be, but when we talk to J2 and John and Davidson, the stuff they're doing with the data, that's disruptive for Cisco. Now, is it going to be like off the charts in terms of a hot startup? No, but for Cisco, that takes something that's in their proprietary or asset base and refactored with AI to leverage that data to provide value to their customers. It could be about documentation. Is it sexy and moving the needle? Yes, sexy, not sexy maybe, but moves the needle for value. That's just value creation. So, I mean, can, is, is, is Gen AI, can Gen AI be evolutionary for some companies? Yes. And I would say yes. Yes, absolutely. Like SaaS companies are going to apply Gen AI and that's going to evolve their platforms, but it's not, I don't see that necessarily as a disruptor. Gen AI is a disruptor. Yeah. But companies like Snowflake are going to take advantage of the disruption because they're going to bring AI to their data. They're not the disruptor in Gen AI. Does that make sense? Do you buy that? I, it's, it's, it's, two things are happening. Sometimes I do buy it, but there's, there's nuance. You can have a disruptive market, market, like with AI. That's disruptive and radically revolutionary. I would say that that's what's happening. I'd consider it a revolutionary market. Sure, I agree. Okay. So, so that's, now the impact of that revolution is the evolution in markets. So let's just take an incumbent, like take Dell or Cisco. They're in mature markets. They've installed bases. They're the incumbents in some markets, but doesn't mean they can't be innovative. That's what I'm saying. I think AI is going to make those guys more nimble actually. It's the guys in the middle who are going to get screwed because they're stuck with, they don't move as fast as they say the startups and the younger companies who are going to get instant power with AI. So you're going to have the rich get richer and be faster so they can leverage their base of their core asset and then get net new capabilities is going to get richer. The startups going to instantly get value and small business will get more value with AI. And then they'll take territory from the mid-range company. So if you're stuck in the mid, as a mid-sized company, you're trying to get territory compete with the big guys. Now you got people nipping at your heels, but you have the, the install base incumbent factor of having an existing business and inertia. You keep, it's hard to move super fast when you're in the middle as a mid-range company. And so, and so, so that they don't move fast with AI they're not going to get it. I think the big guys get bigger and the little guys get faster, use the speed to take territory from the mid-range then the middle gets screwed. Okay. That's my prediction. And we'll look, we'll look back in 10 years and we'll say, yep, that's the pod. You got it right. You got it right again. This is good. You and I argue. And then another analyst will say they invented it. So I coined that term five years ago. You and I argue about this stuff and then we kind of come to a conclusion. So let me just play on this for a bit. Who is the disruptor in generative AI? Open AI. Name the company. Open AI. I wouldn't say open AI and Microsoft. Open AI, you can literally go down the power of law. Wait, wait, wait. Open AI and Microsoft. We agree, correct? No, Microsoft's not the disruptor. They're the financier. They're the ones who are funding it. Okay, but they're partners. They're partners. Okay, they're like the guys who have owned the infrastructure. They are not the disruptors. Is NVIDIA disruptor? I think NVIDIA is enabling that disruption. I think Anthropic, Coher, those are all DNA from the whole Google paper and open AI. They're all ex-open AI people. Open source, open source, the Lama meta is a disruptor by contributing all that code. And then the series of small specialty models are disrupting. And then the entrepreneurs who are actually cranking, this is why the KubeCon stories relevant because those guys are creating the smaller language models. They're the ones doing the test and all the retrieval stuff is disruptive. The vector database has been around for a long time but that as an ingredient in this play is disruptive. So it's a lot of combination of things coming together. I would say that you can even say Amazon's infrastructure as a scalable set of hardware is disruptive. Okay, was AWS the disruptor for cloud? Yes, right, you can agree. They soften the ground. Microsoft created a vehicle funding open AI. AWS created the disruption in cloud, right? Yes, absolutely. Wasn't Netscape a disruptor? Yes, they were there, they moved the needle, they created the trend. Exactly, was Digital Equipment Corporation, which nobody on this podcast has ever heard of, were they a disruptor? Yes, of course. My point is just the disruptor in and of itself, disruption in and of itself does not necessarily equate to business value and success in economic value. Of course, yeah, that's why company. That's why in my innovative score, Dave, ecosystem growth, adoption rate, does it convert? Netscape couldn't convert. Give me the rubric, give me the rubric again. Disrupt. Give me the rubric. What are the factors? Okay, there's many factors. The innovation score that I'm putting together is about looking at a couple like IBM and others who are putting out good stuff and you don't know what's real, right? So I have my own little formula. Is it disruptive? Force, does it have a factor? How much is it factoring into the current market? Will their stuff blend in and ride the wave or be part of the disruption they can play either way? Some companies will ride that disruption wave. Some will participate or do both. Number two, adoption rate of like the technology. Is the market accepting it? Yes or no? Evaluation by the ultimate customer would be, do they develop a ecosystem around that? Partner network, customers, sometimes it's partners indirect and direct selling relationships and integration towards customer features or customer action. And then are they continuing to improve in the R&D of that? Okay, and then little things like, they build a new stuff, how do they have patent to technology? Do they have a moat? Is there longevity to it? Is sustainable, is it durable? Those are kind of the high level. And then there's like a couple of other things in there that are proprietary that look, I don't want to reveal. So the moat might be like a cost advantage, for example. Yeah, library, I mean IBM's got the patent library. I guarantee you they probably have a huge patent library on this stuff. So that would factor in because now they can sustain and protect it and develop around it and use that asset to be a force for good in the marketplace from a disruption and innovation standpoint. So innovation is, what does that do? Do they have the X factor? In this case, it's Watson pun intended. And looks like it's going to ride the current market. I just got to, what's the adoption rate look like? To your point. I have some non-NDA slides I can share with you and I want you to take a look at them. And then you look at them and then other index, other things I'm looking at is like, okay, how is it really changing the customer game? Are there metrics? Are there transformation examples? To what degree is it compatible with the picks and shovels coming in the market? Does IBM align with open source? That's a good question. No idea. I would have asked that question. A lot of the analysts mean it. I would have said, does the red hat align with open source? I mean, what do you mean? Yeah, of course. Well, IBM owns red hat. 100% outright. Of course. And they're doing a lot of AI to make their OpenShift product more protected. It's funny, I'm sitting here like, I know, right? I'm sitting here like defending IBM. I mean, I've been such a critic, watching IBM just destroy CEO after CEO destroying company value for decades. But I'm telling you, John, I think Arvin's got it right. He's just simplified everything and he's brought this. He's got a Nadella-like mentality. Yeah, I think that their big play is to be the operating system for the internet. And I told that to the red hat guys. And I already saw they wouldn't say anything on theCUBE because they're all on a tight lip because it's confidential. But I was kind of squinting through some of the answers. And it's pretty obvious what they're excited about with Edge and all this new use cases is that, you know, multi-cloud and super cloud and distributed computing is basically needs to run on something. And if you look at what they're doing with OpenShift and in KubeCon community backstage, for instance, which is an open source project contributed by Spotify, it's kick ass, all right? And if red hat can be that, you know, put everything into the red hat fold, trusted runtime environment for the internet, so it should be cloud computing, then the clients could get some great AI to be a single pane of glass. Why not have a chat, a GPT that's like your IT department? Okay, automate the knobs and, you know, make sure that I'm standing up data pipelines and LA analytics and all my, there's no zero day threats. Protect my LLMs. That's all platform engineering work that's going on. And so there's a significant shift happening in the marketplace radical right now on this piece. It's huge. There's a security paradigm shift, right? Open SSF, okay, on the threat detection. I mean, there's a lot of data protection products out there, but not a lot on the threat side, Dave. So you got the Kubernetes conundrum, the generative AI dominance, wearables, ethics, you know, future of work, DDoS attacks. I don't know if you noticed this, but OpenAI was down this week because of massive DDoS attacks. I know I was trying to get on and I was like, what's going on? Okay. What use cases do end-to-end stack management? I mean, there's like a serious tech shift happening right now. And a lot of companies like IBM could have great messaging, but if they're on the wrong side of the adoption and the metrics for customers implementing the new architectures for the future, they could have just a great story. And I think, but I do like, Arvin's got the right thing about hybrid and edge. That's not cloud focused. That's not Amazon, that's not Azure. It's mostly future focused. Here's the thing broadly, and I'll probably talk too much about IBM, but I will say this, when Lou Gersner decided to buy PWC and integrate IBM instead of splitting it up into different product divisions, he made a decision which saved IBM, but it turned IBM into a services company. They are a services company first, and they used to be a product company. And Palmasano continued that legacy and Ginny continued that legacy further and Arvin is getting back, in my opinion, to IBM's product roots. And that, to me, is exciting and I think it's important for the United States to have an IBM that does core research, like real research, not just incremental product development, but core research that invents shit. That to me is really, really important because it's, look at Xerox PARC, where's that? Bell Labs, Nokia owns Bell Labs. So those types of core research organizations are few and far between and they're American gems. And I'm hopeful, just like I'm hopeful that Intel will not go bankrupt. Well, Dave, a lot of good conversations there. And again, I think that we should have more of these conversations around what do we think success looks like? Because I'm telling you, with this Genevieve iThane, we're gonna have a huge challenge in this world to figure out what's real and what's not. I like your innovation score, Joan. I like where you're going with this. Well, you're gonna love the cube I'm putting it into. So, because when you want to simplify it, basically imagine having a cube format, Dave, where they double click on an interactive pop-up cube and says, oh, three-dimensional format of ranking companies. Three vectors intersect. So I'm playing with the idea of using the cube data to just have a fun way to kind of keep track of our aggregation and our AI to give a visual representation of making sense of things like that are highly nuanced, but important. Because there's a lot of quality problems out there today, Dave. I mean, I'm seeing people get on TV and saying things that are completely wrong about these companies and they're analysts and they just don't know what they're talking about. And a lot of other analysts are saying, this is bad. And so there's an influencer culture coming around where you can be all fun and throw some haymakers around. We do that all the time just to kind of have fun. But when you're in a service business and you're an influencer or you're an analyst and you represent something that's actually not true, that's not cool. And that's what you're gonna see a lot of that with deep fakes. You're gonna see a lot of that with hot takes. You're gonna see people reparenting back. You got a Tedwind, Dalewind, top line growth. That's not analysis, right? So I think if we have a huge data network, we're gonna bring that together. And I mean, what do companies care about, Dave? What do CEOs care about? I'll answer that. My frustration with like the financial news in particular is they just blow with the wind, right? Oh, the market's down. Also, tech is bad. Oh, the market's up. Tech, that's great. Tech is great. It's gonna save the world. I love Snowflake at 280. Oh, geez, I hate Snowflake at 140. I mean, it just, it changes with the wind. So what CEOs care about is building long-term value. They don't, I mean, if a CEO is worried about his or her day-to-day stock price, my opinion, they shouldn't be CEO. What they should worry about is the job is to deliver shareholder value to expand their TAM. And the way they do that is to make sure they're delivering customer value. They've got strong engineering. They've got great go-to market and their strategy allows them to build a platform that can get them into new markets, but it's all about the long-term value. Unless you want to get sold, clean it up and flip it, do a pump and dump. But the best CEOs that I've observed over time, like the Larry Ellison's, the Michael Dell's, I think Jassy's in it for the long-term. Nadella doesn't care about his stock price day-to-day. I mean, that's just absurd. Why would he? You think Warren Buffett cares about his stock price, what it's doing today. He probably doesn't even look. He probably does look, but doesn't lose sleep over it, I guarantee. So that's what it's all about. And I think that's the best analysts, the best analysts, Wall Street analysts and industry analysts are fundamental thinkers. Dave, you've been an analyst in the industry, the modern era, we're building that out. And most people will look at the changes we're going through in this new environment and where organic community matter, especially with open source and open source content and what we do. What is your vision for the CUBE research? And for the folks watching, we are branding Wikibon as the CUBE research team. And we'll have a site for that. We're going to continue to do more and publish more of our research, continue to make things as fast as possible. You'll see a lot more videos, just a lot more AI hitting the screen. I'm loving the CUBE AI, both on the CUBE's video ingestion as well as the knowledge graphs we got an expert network, 18,000 people, 35,000 videos. Dave, what is the CUBE research vision? I mean, as a modern era, as the world changes. Yeah, so customers need new service, a new kind of service. Is the game the same? Does it remain the same? Does it change? What's your vision? Well, I mean, you know, I mean, we're all about, you know, changing John and speaking of disruption, that's us. We have had a front row seat for over a decade with the CUBE on high quality content, how high quality content is created, distributed, consumed by audiences, what audiences care about and how they get value out of content. And so the CUBE research is all about contributing to the silicon angle and CUBE vision of the CUBE global and our global community and bringing a knowledge base that is fact-based, that's data-based, that can back up our opinions and serve both our audience, which is practitioners and elite business technology people and can serve our customers, our clients, who really want to get their story in front of that audience with content that matters. And that's really what the CUBE research is all about. I'll share a quick story for a little bit of a prop but I was going to, I was in Chicago, a public company CEO came up to me and said, you know, the partnership with the CUBE and the silicon angle will keep on. Now the CUBE research advisory has been a game changer for us, the partnership approach. Partnering has been, was a key point there. And then I ran it to a head of analyst relations from another company and she said that you guys go beyond just helping craft, you know, the stories and positioning and forecast but you guys help us build the future, okay? And you guys don't even have a retainer or you guys should do that immediately. You guys help our stakeholders. And then, and then finally the CMO said, you guys help us lead the competition with the content and seeing the trends and helping the content strategies map to augmenting their role. And I think this modern content market's coming where the website's going to be just, you know, infrastructure, you're seeing that with the videos now and the AI being like the co-pilot. I think that's where I see the best intersection is just having that trusted relationship with our customers and allowing us to do great content. Listen, if you're a company and you're not completely rethinking your entire content strategy and how to reinvent AR and PR, you're going to be left behind. I mean, how you're using real time data, obviously AI and in the moment knowledge, like it's a speed game, but it's also a quality game. And the other thing too, John, is community. I was thrilled at the IBM facility. So myself, Sanjeev Mohan, Tony Bear and Merv Adrian did a collaboration because that's, we're very collaborative. We're collaborative with other analysts. You know, we're not the only analyst firm in the world, right? I mean, there's obviously Gartner, IDC. If any analyst does good work, I want to learn from it because I have great respect for other analysts. IBM this week, man, there were some smart people. They were like David Floyer class people going deep. And I'm like, wow, that individual is smart. I want to learn from her or I want to learn from him. And they are welcome in our community and we'd love to amplify their research. We've worked with every type of analyst firm out there from Gartner, IDC, Forester, ESG, you know, you name it. We've opened our arms to have people. I'll tell you, I've been doing a lot of talking with heads of comms and heads of ARs and top CMOs. And what's happened has been some big blowback lately and on a couple of examples. And this is going to be a first generation problem as you get more content into the analyst game where it's organic. Traditionally an earned media play or organic content like what we're doing and some of the services we do as a trusted partner, for example, credibility is everything. And once someone does something in the community that's not good for the reputation, the client gets the blowback. So you're starting to see this on Instagram day where these influencers will say whatever it takes to get the product sold. And what's happening is CMOs and executives saying reputation risk is out there. So what's going to happen in B2B is you want to be judged by the company that you keep. And the analysts that you choose, if they're out there throwing fake information around and not having the depth, that's on you. So I think you're going to see a couple of blowbacks come back on that, reputation is everything and having substance will matter. And you know, the one thing of doing the Q for 13 years is, you know, we've seen the movie a few times and you know, we've had the long game approach of just hanging out with great people, talking to smart people and getting the trends right and being part of a great service. And so a little bit of a rant here on the market of how it's changing, CMOs, head of comms, head of analyst relations, this is the new marketing formula for the companies because everyone is now seeing everything. So you can't hide behind fake news, fake analysis, fake influence, you got to show it on a consistent basis. And that's going to be the new filter, Dave, I think, you're going to see that. So I'm excited to see what future can bring with theCUBE research and build a net bridge to the future. Well, and I want to shout out to our colleagues at ETR. I've had a multi-year, well over three years now partnership with ETR predates the pandemic. And the more I get to know them, the more high-quality people who do amazing data work, they're not just a survey house, right? I mean, these guys, they truly are data analysts. They've got real data scientists and they're just high-quality people of impeccable ethics. I saw Ray Wang, I saw Sanjeev and I was a bunch of great analysts out there. You know what the common theme is when I talk to folks, it's a service business and he's probably a good service. Ray Wang, man, he was at IBM and he was at Dell. He's so good. I mean, at dinner last night, you know how good he is on theCUBE. Yeah, yeah. And Ray is just, the guy's so knowledgeable about what's going on in the world, what's going on in tech, and he's just super articulate. I really like him a lot. Well, our rant section turned out to be a rant on the media itself, Dave, and the analyst relations PR CMO equation. I am super excited that the media business is going to shift in-house to these companies. They're going to have needs and service and I think we can help them out there. We've got big events coming up. I'll be at super computing next week. We've got coverage of Microsoft Ignite. We also have Reinvent coming up. We're going to have a SuperCloud 5 special edition in studio in Palo Alto. We're going to have the Lama Index CEO, a bunch of top AI folks in the Valley coming into Palo Alto. We'll have a live stream for two days, Lisa Martin, Savannah Peters, and we'll be hosting in studio. You and I will be in Vegas for AWS's annual conference. We're going to talk to all the top people, do an interviews, check out that stream, and then, you know, end of the year. Dave, can't believe it. Well, that's going to be exciting, John. I mean, you and I are going to be on the ground. We're just, we have multi-continents going on here, right? Rob and Rebecca out in Barcelona, HPE Discover. Yeah, SuperCloud 5. Five SuperClouds, John. Four SuperClouds this year. AI madness. Well, AI madness is just so much fun. I got to say, I love this market. I love what's going on. I love the disruption. I love how everybody's winning. We're going to watch this mid-range. I might not be accurate on that, but my gut's telling me that the big winners will be the big players and the startups and small companies. So, you know, right now, that's who I see getting the most value of AI. So this is just the beginning, you know, inference is the new web app, as Tim Hawkins at Google said on stage at KubeCon. I totally love that. I think he's right on the money. I think you, this next generation of developers in open source are going to set the agenda. And you watch, you watch the big guys get dressed for the party. Okay, Dave, but you, the action is in open source. And that's like where everyone's trying to read the tea leaves. So if you're not an open source and you might miss what's going on there, great benchmarks, great technology, great innovators. And then the big guys are tooling up for the picks and shovels and for the market. It's going to be huge. All right. Dave. Hi, John. Have a great weekend. Yeah, I'll talk to you probably tomorrow. That's a wrap for 37 Pod. See you next time.